现代电子技术2016，Vol.39Issue(13)：63-66,71,5.DOI:10.16652/j.issn.1004-373x.2016.13.016
基于卷积神经网络LeNet-5的货运列车车号识别研究
Research on freight train license recognition based on convolutional neural network LeNet-5
摘要
Abstract
For the character recognition of freight train license,the improved recognition method based on convolutional neu⁃ral network LeNet⁃5 is proposed. Considering the structural features of the hierarchical convolutional neural network and local field,the parameters of quantity and size of each layer feature pattern in the network were improved correspondingly to form the new network model suitable for the freight train license recognition. The experimental results show that the proposed method has strong robustness to solve the license breakage and stain,and high recognition rate,which provides a guarantee for the accuracy of the entire license recognition system.关键词
列车车号/车号识别/卷积神经网络/LeNet-5Key words
train license/license recognition/convolutional neural network/LeNet-5分类
信息技术与安全科学引用本文复制引用
王晓锋,马钟..基于卷积神经网络LeNet-5的货运列车车号识别研究[J].现代电子技术,2016,39(13):63-66,71,5.基金项目
国家自然科学基金（61171156）支持项目 （）
